ADDENDUM 
It was impossible to insert in the text the results of the writer’s 
field work in the summer, autumn, and winter of 1912, but the most 
important results may be summarized briefly here. 
Piegan (of Montana) has whispered vowels terminally after w and 
nasals; x is distinctly post-velar; final & is distinctly aspirated. Gros 
Ventre (Atsina) sheds little light on Arapaho, sharing with the latter 
practically all deviations from normal Algonquian. Potawatomi dif- 
fers more from Ojibwa, Ottawa, and Algonkin than these do from 
one another. According to communications from Doctor Sapir of the 
Geological Survey of Canada and Doctor Radin of the International 
School of Ethnology and Archeology, the Ojibwa dialect at Sarnia, 
Ontario, seems to be highly specialized. The intimation given in the 
section on Delaware that Zeisberger’s material represents no single 
dialect was borne out by the writer’s experience with the Munsee of 
Kansas and the Delaware of Oklahoma. Apparently no distinction 
can be drawn to-day between Unami and Unalachtigo. The pho- 
netic system of Zeisberger is very deficient. Every stop occurs as 
surd, sonant (after nasals), surd aspirate (terminally), and glottalized. 
Voiceless ? occurs medially before consonants in both Delaware and 
Munsee, and terminally in the latter (where it seemingly is lost in the 
former). Long consonants are common, also consonantic clusters, 
owing largely to elimination of vowels. Umlaut is caused byw. On 
the whole, both Delaware and Munsee have suffered very considerable 
phonetic changes from normal Central Algonquian; Munsee is by far 
the more archaic of the two. In Munsee whispered vowels occur 
initially, medially, and terminally (after w). In Delaware seemingly 
they are found medially and terminally after w. In both, s, y, w, and l 
occur glottalized as well. The variety of forms given in the table is 
due in part to dialect mixture, in part to phonetic changes. Some of 
the forms are due possibly to mishearing; some contain double 
objects; others seemingly are to distinguish third persons; still others 
owe their origin to causes which are unknown although the forms 
exist to-day. The statement that one dialect had the closest rela- 
tions with Cree-Montagnais and another with Shawnee, is wrong. 
Zeisberger’s inadequate phonetics were wrongly interpreted. It is 
clear that both Delaware and Munsee are closely related and, though 
they can not be easily classed with any other large group, it is clear 
that they approximate the Ojibwa group in important points, and 
290a 
